Meta Scraps Project OT After AI Agents Fail to Deliver, Cancels 60% Team Cuts Plan

Reviewed byNidhi Govil

15 Sources

Share

Meta abandoned its ambitious Project OT plan to become AI-native after AI agents failed to boost productivity as expected. The company explored slashing team headcounts by 60% and replacing human roles with AI agents, but canceled the second wave of layoffs hours before the first, following employee backlash and internal data showing AI produced more code but fewer features.

Meta's Bold Vision for AI-Driven Organizational Transformation

Meta explored one of tech's most aggressive AI-focused restructuring plans this year, code-named Project OT (Organization Transformation), which would have slashed some team headcounts by as much as 60 percent

1

. Mark Zuckerberg and his top executives hatched the plan during their January leadership retreat at his Hawaii compound, envisioning an AI-native organization where AI agents would take over much of the daily work performed by thousands of human employees

5

. The restructuring would be carried out in two waves, with the first purge scheduled for May and a second shake-up planned for November

5

. One HR executive projected the workforce reduction would be at least as large as Meta's 2023 cuts of roughly 25 percent or more

1

3

.

Source: Reuters

Source: Reuters

The AI-Native Playbook and Organizational Vision

Zuckerberg's plan to replace staff with AI centered on creating smaller, "talent-dense" cadres of human staffers who would oversee virtual workers

5

. An internal document titled the "AI-Native Playbook" outlined how product teams of 10 to 20 specialists would shrink to pods of three to five people

4

. Engineers, designers, and product managers would transition into general-purpose "builder" roles, while middle-management layers would shrink

3

. The vision described an AI-native company where "AI-ready tools and agents interact, workflows are automated, [and] new builds are AI-first"

1

. Agent-assisted analysis would help prioritize daily tasks, and the company would sell AI agents to third parties, which Meta began doing in June

1

.

Source: Ars Technica

Source: Ars Technica

When AI Agents Failed to Deliver Productivity Gains

Meta's own internal data revealed the critical flaw in Project OT: AI agents were producing volume rather than value. Code changes to internal software platforms and infrastructure surged 220 percent year-over-year, according to a June post by CTO Andrew Bosworth, but changes that led to new or upgraded features reaching Meta users increased only 36 percent

1

4

. The gap between code production and actual feature deployment exposed fundamental limitations in the AI-driven organizational transformation strategy. Internal posts revealed that unchecked AI agents were performing "large-scale, disruptive actions that humans are unlikely to execute," leading to a 40 percent increase in major technical and security incidents compared to the prior year

1

4

. Employee time spent resolving these problems increased by as much as 70 percent

1

.

Employee Backlash and Morale Crisis

Meta AI restructuring efforts triggered significant employee backlash as workers became convinced they were training their replacements. The company had reassigned engineers to write software puzzles used as training data, and mandated tracking software on US employees' devices to capture keystrokes and mouse clicks so AI agents could learn to use computers like humans

4

. Meta paused that program in June amid mounting concerns

4

. Reports of impending layoffs and the tracking program severely damaged morale

1

. Meta's internal sentiment score plummeted from 74 percent favorable to 55 percent in its half-year Pulse survey

4

. Employees responded to executive posts with pictures of elephants, and labor organizing gathered pace

4

. Twenty-six employees have since sued the company, alleging its AI systems targeted workers on medical leave for layoffs

4

.

Zuckerberg's Abrupt Course Correction

On the night of May 19, just hours before the first wave of layoffs, Zuckerberg canceled planning for the November cuts

5

4

. Meta proceeded with laying off approximately 8,000 jobs the next day, roughly 10 percent of its workforce, and moved 7,000 people into AI roles in the same week

4

. Reuters was unable to determine what exactly prompted Zuckerberg to shift course

1

5

. In July, Zuckerberg acknowledged during a company meeting that the "trajectory of the agentic development over at least the last four months hasn't really accelerated in the way that we expected"

1

.

Source: NYMag

Source: NYMag

The Financial Reality Behind AI-Native Ambitions

While Meta explored slashing team headcounts to reduce its towering $30 billion worker compensation budget in the first half of the year, the company faced mounting capital expenditure on data centers

2

. Investment is estimated to hit nearly $170 billion next year, according to LSEG data

2

. At that pace, increased depreciation and amortization might well consume any employment savings from cost-cutting measures. Meta's total employee compensation is actually rising—ignoring severance costs, the company spent nearly 30 percent more in the first half of 2026 than in 2025

2

. AI researchers command premium salaries, and Zuckerberg has notoriously bid up compensation as much as anyone

2

.

What This Means for AI Adoption Across Tech

Meta's stumble with Project OT highlights fundamental challenges organizations face when determining the best uses for AI agents and when the technology truly outperforms human employees. As Clayton Christensen's "The Innovator's Dilemma" explains, existing products, employees, and managerial structures are well adapted to the status quo, and attempting to change them in the face of revolutionary industry shifts might cause them to fall apart

2

. Moving fast and breaking things works for startups with little to lose, but a $1.4 trillion titan like Meta cannot afford that mindset

2

. The episode demonstrates that drastically refashioning a giant firm presents inherent problems, and that it's rarely evident how best to use new technologies. Factories and offices took decades to optimally redesign work around electricity and computers

2

. Watch for how Meta and other tech giants balance AI automation ambitions against employee concerns, technical limitations, and the reality that sophisticated models may not yet be ready to replace human roles with AI agents at the scale once envisioned.

Today's Top Stories

© 2026 TheOutpost.AI All rights reserved